Three metals are best for sensitive ears: S316L Surgical Steel, pure 24 karat gold, and implant-grade titanium. All three are stable, meaning they resist corrosion and therefore release very little into skin. But the more useful answer is that a named grade of any of them beats an unnamed version of all of them, because you can't predict what you can't identify. Our materials page names ours exactly.
Why stability is the whole criterion
Skin doesn't respond to a metal's name, price or reputation. It responds to what that metal releases into it.
Every metal releases something in the warm, damp, mildly acidic environment of a pierced ear. What varies enormously is how much, and that rate is governed by corrosion resistance.
A metal that resists corrosion holds its constituents rather than shedding them. A metal that corrodes liberates ions into the tissue around it. That's the entire mechanism behind green marks, itching, and the exposure that leads to sensitization.
So "best for sensitive ears" translates to "most stable, and identifiable enough that you can be confident which one you have."
The three metals that are best for sensitive ears
| S316L Surgical Steel | Pure 24 karat gold | Implant-grade titanium | |
|---|---|---|---|
| Contains nickel | Yes, 10 to 14 percent | No | No |
| Nickel release | Commonly below detection under EN 1811 | Nothing to release | Nothing to release |
| Corrosion resistance | Excellent | Excellent, chemically inert | Excellent |
| Tarnishes | No | No | No |
| Weight | Light at stud sizes | Light | Very light |
| Hardness | Good | Soft, so used as a finish | Good |
| Cost | Lower | Higher | Higher |
| Fine settings possible | Yes | Yes, over a core | Harder |
S316L Surgical Steel
Low carbon, with molybdenum added. The low carbon reduces the chromium carbides that create weak points where corrosion begins. The molybdenum substantially improves resistance to chlorides, which are present in perspiration and pool water.
It's the grade used for medical instruments and surgical implants, chosen for exactly the property that matters here.
The nickel question. It contains nickel, as almost all austenitic stainless steels do, because nickel is what stabilizes that structure. Presence and release are different measurements though, and only release affects skin. EN 1811 is the test.
Pure 24 karat gold
All 24 parts are gold, with no alloy at all. Nothing present to react. Gold is one of the least reactive metals there is, which is why it's used in dentistry and electronics.
The catch. It's soft, so it works as a finish over a stronger core rather than as a whole earring. Which makes the core the real question, and it should be a named stable metal rather than an unnamed base.
Implant-grade titanium
Genuinely inert, contains no nickel, and very light.
The catch. "Titanium" alone is as uninformative as "stainless steel" alone. Grades vary, and common alloys like Ti-6Al-4V contain aluminum and vanadium. It's also harder to work, so fine settings and design range are more limited.
What disqualifies the rest
9K to 14K gold. Between 10 and 15 parts in every 24 are alloy, and the recipe is almost never published. Commonly contains copper, sometimes nickel.
White gold. Gold is yellow, so whitening it requires an additive. Nickel does it cheaply; palladium does it expensively. If nobody will say which, treat that as unanswered.
Rose gold. The pink color is copper. There's no version without it, at any karat, from any brand.
Sterling silver. 925 means 92.5 percent silver. The remaining 7.5 percent is usually copper, which tarnishes and turns skin green.
Plated anything over an unnamed base. Fine while the coating holds. Every coating wears through, first at the post, and then the base metal is what's in contact.
Unnamed "stainless steel." Could be 316L, could be 430. The family includes hundreds of alloys with quite different behavior.
Brass, nickel silver, pot metal and unnamed base metals. Nickel silver contains no silver.
The distinction that resolves most arguments
Presence is a question about composition. Release is a question about behavior.
Nickel bound tightly inside a stable structure with an intact protective oxide film releases almost nothing measurable. The same element in an unstable alloy, or in a piece that's corroding, releases considerably more.
Which is why a well-specified steel containing 12 percent nickel can genuinely be a better choice than something marketed as nickel free whose composition nobody has verified.
EN 1811 measures release: a week in artificial sweat at body temperature, then analysis of what came out. It's the only number in this whole subject that describes an actual product rather than a category.
Why the grade matters more than the metal
This is the part that changes how you shop.
A named 316L beats an unnamed titanium. A stated 24 karat over a stated steel core beats an unstated "gold". The reason is simple: you can look up a specification and you can't look up an adjective.
So the practical ranking isn't titanium first, then gold, then steel. It is:
- Any of the three, with a grade stated and a test result published
- Any of the three, with a grade stated
- Everything else
Most of the category sits in the third tier, and most of it isn't dangerous. It's just unknowable, and you'd be finding out on her ears.
Why so many listings can't tell you
Not usually deception. Structure.
A retailer buys from a distributor who buys from a factory. Each step passes along the description it received. By the time a listing is written, the person writing it is several removes from anyone who could answer a metallurgical question, and a category name is the only description that survived.
I've dealt with that chain directly. When I was sourcing, the same specification questions produced answers that didn't match each other between suppliers. Some of what was described one way was clearly something else. There was very little malice in it. Nobody in the chain was positioned to check.
That's why I turned down suppliers rather than accept a specification I couldn't verify. There's considerably more available on the market in styles and options than we carry. Taking it would have meant losing the ability to answer the only question that matters.
Why this matters more for a child
Nickel allergy affects around one in five people in North America. The American Academy of Pediatrics reported in 2020 that nickel allergic contact dermatitis in children has nearly quadrupled over three decades, with at least 1.1 million children in the United States already sensitized. A 2025 systematic review found people with piercings were nearly six times more likely to be nickel allergic.
Sensitization is cumulative and generally permanent. The immune system reacts once it's been exposed enough, over enough time, to learn the metal.
Childhood is when that exposure begins, which is why the everyday earring matters more than any single special pair. Most of her lifetime total comes from whatever she wears without thinking about it.
Worth knowing: the United States has no legal limit on nickel release from jewelry. The EU introduced one in 1994 and tightened it in 2004. Denmark restricted it in 1992, and nickel allergy among Danish children fell from 24.8 percent to 9.2 percent between 1985 and 1998.
The component nobody checks
Whichever metal you choose, ask one more question: are the post and the back the same material as the front?
Backs are commonly generic components, bought in from a different supplier than the earrings themselves. The manufacturing conditions are unknown, the finish may not match, and where testing exists it may not have covered the back at all.
The post matters even more, because it sits inside the piercing channel where skin is thinner, warmer and more permeable than on the outside of the lobe.
A brand that can answer this knows its own product. Most can't.
How to compare two listings in practice
Reading about metals in the abstract is easy. Applying it to two open browser tabs is where people get stuck, so here is the sequence.
Find the numbers. Search each page for a digit attached to a metal. S316L, 316L, 24K, 18K, 925, Grade 23. A page with metals but no numbers is describing appearance rather than composition, and that is your first sort.
Find the components. Search for the words post and back. If neither appears, the description almost certainly covers the visible part only, and the components with the most skin contact are unaccounted for.
Find the tester. A laboratory name means a third party checked. The word tested on its own means somebody intends you to feel reassured.
Find the result. Not a badge, not a phrase, an outcome. Nickel release not detected reads differently from meets all safety standards, and the difference is that one of them can be wrong in a way you could discover.
Then compare prices. Not before. Price sorted first will lead you to the wrong tab almost every time, because price tracks design and brand rather than specification.
Two listings that both pass the first four are a genuine choice between good options. Two that both fail are not a choice at all, they are two unknowns, and the cheaper unknown is not better value than the dearer one.
What we chose
purepixxi uses two of the three, and the reasoning is worth stating.
S316L Surgical Steel, named in full every time. On the Polished Surgical Steel range it's the entire earring, one material throughout, with no coating and no second supplier.
24K Pure Gold, which is pure 24 karat gold finished onto a solid S316L Surgical Steel core with Aunix™ 24K, our engineered gold finish. It's a gold-plating process and we say so. All gold plating is only as safe as the metal underneath, which is why we tell you what ours is.
Not titanium, for two reasons. The EN 1811 result answers the nickel question with a measurement on our actual product rather than an inference from a composition table, and titanium would narrow the fine settings considerably.
Not white gold, rose gold or silver, because all three require metals we won't put in.
On both ranges the setting, the post and the butterfly back carry the same S316L Surgical Steel core.
Independently tested by SGS. I chose the pieces with the largest surface area touching an ear and had all three components tested, across eight pairs. Nickel release not detected under EN 1811. Tested to ASTM F2923 and CPSIA, with lead below a tenth of the United States legal limit. Report BTO 341803/1, and you can ask me for it.
You can see the range on the Polished Surgical Steel collection.
Questions parents ask next
Is platinum an option? It's stable and inert, and it's expensive and uncommon in children's jewelry. The same rule applies: the specification, not the name.
What about niobium? Genuinely inert and nickel free, widely used in body jewelry, and rare in children's designs.
Is medical-standard the same as implant-standard? No, and the words get used loosely. What matters is a grade designation you can look up, not a descriptive phrase.
If she reacted to steel before, does that rule out S316L? Not necessarily. Most "steel" earrings aren't 316L, and the reaction may have been to a lower grade, a plating, or a back from a different supplier.
Does the stone matter? The stone doesn't touch her skin in a stud setting. The metal it's set in does.
If you would rather not have to ask any of this again, the 4mm Flawless Crystal Stud Earrings in Surgical Steel are S316L throughout, with nothing plated over anything.
Related reading
- What earrings are best for toddlers?
- What earrings can she wear every day?
- What earrings won't irritate her ears?
- What children's earrings are actually made of
About the author. Matt Westmore is the founder of purepixxi. He came to children's jewelry from manufacturing and sourcing rather than marketing, which means he has walked the factory floors and knows what is actually inside a product rather than what a label says about it. purepixxi uses two materials only, names them exactly, and has them independently tested.
